sign in sign up
<<
Home , Malolactic fermentation, Yeast in winemaking

foods

Article Impact of the Timing and Temperature of Malolactic on the Aroma Composition and Mouthfeel Properties of

Anthony Sereni, Quynh Phan, James Osborne and Elizabeth Tomasino *

Department of Food Science & Technology, Oregon State University, Corvallis, OR 97331, USA; [email protected] (A.S.); [email protected] (Q.P.); [email protected] (J.O.) * Correspondence: [email protected]; Tel.: +1-541-737-4866

 Received: 13 May 2020; Accepted: 15 June 2020; Published: 18 June 2020 

Abstract: (MLF) is an important process in wine production due to the resulting reduction in acidity. MLF is typically induced by the addition of after the completion of alcoholic fermentation (AF), but can occur concurrent with AF by co-inoculation of O. oeni with cerevisiae. This study investigated the effect of MLF inoculation timing and temperature (15 ◦C and 21 ◦C) and the presence of the non-Saccharomyces Torulaspora delbrueckii on Chardonnay wine aroma and mouthfeel. Aroma composition was measured using headspace solid-phase microextraction-gas chromatography mass spectrometry (HS-SPME-GCMS). Mouthfeel attributes of the produced were assessed by a winemaker panel, using Napping® and Ultra-flash profiling. Significant differences in aroma composition and mouthfeel perception were found based on MLF timing and inoculation conditions, as well as between temperatures. Temperature had a greater impact on the aroma composition for sequential inoculations, while there were little differences based on the temperature of concurrent . Treatment type and temperature also affected the chemical composition of finished wines. Mouthfeel was impacted, although not as strongly as aroma composition. These findings demonstrate the usefulness of various MLF practices to influence the sensory qualities of a Chardonnay wine.

Keywords: ; Oenococcus oeni; Torulaspora delbrueckii; concurrent fermentation; GCMS; Napping®; ultra-flash profiling

1. Introduction Chardonnay is referred to as a neutral aromatic cultivar, producing a wine that is not defined by a specific set of aroma compounds [1,2]. Chardonnay wine is possibly the most diverse style, allowing for many variations in processing steps, including a variety of styles in and some dessert wines. However, it is most commonly used for the production of still white wine [3]. Chardonnay are typically pressed before alcoholic fermentation (AF) to minimize contact with the skin and seeds [1]. In the absence of the aromatic constituents of grape skins, Chardonnay is most often fermented at one of two temperature ranges to retain aroma compounds; 10–16 ◦C or 20–25 ◦C[4]. However, the impact of fermentation temperature on mouthfeel has been little studied. Of almost all white wines, texture and mouthfeel are considered of extreme importance for the Chardonnay wine style [5]. Another important process for many styles of Chardonnay wine is malolactic fermentation (MLF). This process results in a raise of pH in the wine and an increase in microbial stability due to the removal of malic (a potential nutrient source for spoilage bacteria) [6].

Foods 2020, 9, 802; doi:10.3390/foods9060802 www.mdpi.com/journal/foods Foods 2020, 9, 802 2 of 14

Because of the decrease in acidity, this process is conducted in cooler climate cultivars, such as Chardonnay, which generally have higher acidity than warm climate grapes [7]. MLF is generally conducted after AF, with the addition of Oenococcus oeni, for the conversion of diprotic to lactic acid (single protic group). This process is also known to alter the sensory characteristics of wines. Avedovech et al. [8] found that tasters could discern differences in aroma between Chardonnay wines that have undergone MLF vs. non-MLF treatments. Subsequent volatile composition analysis by Gas Chromatography Mass Spectrometry (GCMS) also showed significant differences. MLF produced a variety of desired aroma compounds, including , , volatile , diethyl succinate, volatile , ethyl , n-propanol, 2-, n-hexanol, , and 2,3-butanediol. While MLF is typically conducted after the completion of the alcoholic fermentation, it may also occur at the same time as the alcoholic fermentation (AF). This is known as either concurrent or simultaneous fermentation and is induced by the inoculation of both the yeast and bacterial starter cultures at the same time. In red winemaking, concurrent inoculation has been studied as a possible means of reliably completing AF and MLF in a shorter period of time. For example, AF and MLF were shown to more reliably complete fermentation with concurrent inoculations than sequential AF and MLF [9]. Concurrent AF and MLF is increasingly being used in modern winemaking, but there are still anecdotal concerns of higher levels of volatile acidity and stuck fermentations. This is due to the fact that O. oeni is a heterofermentative bacteria that can produce via the metabolism of [10]. However, due to the bacteria’s preference for malic acid metabolism at pH levels <3.60, increased acetic acid has only been noted when concurrent inoculation occurred in high pH /musts [11]. A number of studies have reported on the impact of concurrent inoculation on wine aroma [12–15]. Unfortunately, few have reported on the influence of MLF timing on wine mouthfeel. For example, Maarman et al. [15] found increases in volatile esters including when using concurrent inoculations versus sequential inoculations. However, sensory analysis was not conducted on these wines, so no comparison of wine mouthfeel could be made. In a study conducted in Shiraz, Abrahamse and Bartowsky [12] reported significant differences in volatile compounds as well as and pigmented polymer composition. Again, no sensory analysis of the wines was conducted, so the influence of MLF timing on wine mouthfeel could not be determined. If concurrent fermentations are to be performed, the presence of other micro-organisms at the beginning of fermentation must be considered. After grapes are harvested and processed, a large number of yeast and bacteria species may still be present on the grapes. How the presence of these microbes influences the ability of O. oeni to conduct a concurrent MLF is relatively unknown. A number of studies have investigated how the presence of naturally present on the grapes at impact Saccharomyces cerevisiae. Many of these yeast species present on grapes cannot survive the high environment created by the fermentation by Saccharomyces cerevisiae, but can still interfere with the health of Saccharomyces cerevisiae by limiting nutrient availability or by the generation of harmful compounds [16,17]. Arnink et al. [18] demonstrated the negative impact of nutrient stress on S. cerevisiae and O. oeni during AF and MLF, with particular importance on nitrogen availability. High microbial loads can also influence the sensory properties of a finished wine in both positive and negative ways [17]. Some non-Saccharomyces yeast species can positively impact wine flavor and aroma when present before or during fermentation with S. cerevisiae. In fact, a number of non-Saccharomyces yeast are now available as commercial cultures. Metschnikowia pulcherrima has been shown to decrease final wine alcohol content from 0.9–1.6%, with positive sensory aspects noted for Shiraz wines, but a negative aromatic influence imparted into Chardonnay wine due to increased levels of ethyl acetate (aroma described as nail polish remover) [19]. Hanseniaspora vineae, an apiculate yeast species, has been cited to remain active in ferment until up to 9% alcohol (v/v) and in Chardonnay wines, it can be associated with an increase in “fruit intensity, described as banana, pear, , citric fruits, and guava.” [20]. Englezoset et al. [21] has shown that different yeast species and inoculation protocol greatly impact Foods 2020, 9, 802 3 of 14 the growth of and the wines’ aroma composition. Despite the importance of non-Saccharomyces yeast to the winemaking practice, little is known about how their presence may impact the MLF. This is of particular importance when considering concurrent fermentations as it is unknown how O. oeni will react when inoculated into a grape /where a high population of non-Saccharomyces may be present. The objective of this study was, therefore, to investigate how the timing of MLF impacts the aroma and mouthfeel of Chardonnay wine. Two different temperatures of fermentation were explored as Chardonnay is commonly fermented at either a cool (13–15 ◦C) or warm (18–21 ◦C) temperature to produce different styles of wine. The impact of a high population of a non-Saccharomyces yeast on co-inoculated fermentations was also investigated as the use of this is increasing, but little is known about how this may impact MLF.

2. Materials and Methods

2.1. Winemaking Chardonnay grapes were harvested September 2014 from Oregon State University’s Woodhall (Monroe, OR, USA). A destemmer (VLS technologies, Treviso, Italy) was used to destem the grapes, which were then pressed at 0.1 MPa for 15 min using a membrane press (Velo technologies, Treviso, Italy). The resulting juice was settled for 12 h at 8 ◦C. After , the juice was divided into 24 one-gallon glass carboys, 3 L per carboy, and secured with airlocks. A commercial culture of Torulaspora delbrueckii (Vinoflora Prelude™) (Chr. Hansen, Hørsholm, Denmark) was added to one set of carboys at a rate of 0.25 g/L after hydration according to the manufacturer’s specification and the juice was held at either 15 or 21 ◦C for 48 h. After 48 h, carboys were inoculated with Saccharomyces cerevisiae D47 (Lallemand, Montreal, Canada) at a rate of 0.25 g/L after hydration according to the manufacturer’s specification. Carboys of juice to which T. delbrueckii was not added were also inoculated with S. cerevisiae D47 after the initial settling step. At the time of S. cerevisiae inoculation, half of the carboys were also inoculated with Oenococcus oeni Beta (Lallemand) to induce MLF. O. oeni was inoculated at approximately 1 106 cfu/mL following the manufacturer’s instructions. For the remaining carboys, × Beta was inoculated at the completion of alcoholic fermentation. All fermentations were performed in triplicate at either 15 or 21 ◦C. Figure1 displays a flow chart of the treatment details. At the completion of AF and MLF (glucose + <4 g/L, malic acid <50 mg/L), an addition of 50 mg/L SO2 was made to the wines before they were placed at 4 ◦C to settle. After 14 days of settling, the wines were racked and free SO2 was adjusted to approximately 25 mg/L. Samples from each replicate were taken and frozen at 20 C until required for analysis. Replicates were then combined, − ◦ sterile filtered (0.45 µm PES cartridge filter), and bottled in 375 mL green glass bottles sparged with nitrogen and sealed with aluminum closures (Stelvin TM, Amcor, Australia). Foods 2020, 9, x FOR PEER REVIEW 4 of 15 Foods 2020, 9, 802 4 of 14

Chardonnay Grapes

Woodhall Vineyard, Monroe, Or

Inoculation with T. delbrueckii

Sequential Sequential inoculation inoculation for for AF, MLF AF, MLF Concurrent inoculation Concurrent of AF and MLF inoculation of AF and MLF

Figure 1. Chardonnay winemaking treatments (AF = alcoholic fermentation with S. cerevisiae D47), MLF = malolactic fermentation with O. oeni beta). Winemaking treatments were done at 15 ◦C and 21 Figure◦C. AF 1. was Chardonnay conducted winemaking by S. cerevisiae treatmentsD47 and MLF (AF by= aO.lcoholic oeni Beta. fermentation with S. cerevisiae D47), MLF = malolactic fermentation with O. oeni beta). Winemaking treatments were done at 15 °C and 21 2.2. Chemical°C. AF was Analysis conducted by S. cerevisiae D47 and MLF by O. oeni Beta. Basic juice analysis included total soluble solids, pH, and titratable acidity (TA). Soluble solids 2.2.were Chemical monitored Analysis throughout AF using a digital densitometer (Anton Paar, Santner Foundation, Graz, Austria).Basic pHjuice was analysis determined included by -selective total soluble electrode solids, pH (ThermoFisher, and titratable Scientific, acidity MA,(TA) USA). Soluble and solids TA by weretitration monitored with 0.1 throughout M NaOH. GlucoseAF using/fructose, a digital malic densitometer acid, and ( aceticAnton acid Paar, were Santn measureder Foundation, by enzymatic Graz, Austria).test kits (r-Biopharm, pH was determined Darmstadt, by ion Germany),-selective while electrode (ThermoFisher was determined Scientific, using an MA, Alcolyzer USA) and (Anton TA byPaar, titration Santner with Foundation, 0.1 M NaOH. Graz, Austria). Glucose/fructose, malic acid, and acetic acid were measured by enzymaticAroma test compounds kits (r-Biopharm, were measured Darmstadt, using Germany HS-SPME-GCMS,), while ethanol adapting was determined methodology using 1 from an Alcolyzerpreviously (Anton published Paar,[ 22Santner]. Samples Foundation, were run Graz, on Austria). a Shimadzu GCMS-QP2010 (Shimadzu Scientific InstrumentsAroma compounds Inc., Columbia, were MD, measured USA) equipped using HS with-SPME a CTC-GCMS Combi-Pal, adapting autosampler methodology fitted 1 with from a previouslystack cooler published (CTC-Analytics [22]. Samples AG, Switzerland). were run on Sample a Shimadzu preparation, GCMS extraction,-QP2010 (Shimadzu fibers, columns, Scientific oven Instrumentsramp, and all Inc., other Columbia, method parameters MD, USA) are equipped the same with as Tomasino a CTC C etombi al. [-22Pal]. autosampler Samples were fitted analyzed with on a stacka Shimazdu cooler (CTC QP2010-Analytics GCMS (Shimazdu,AG, Switzerland). Columbia, Sample MD, preparation, USA) equipped extraction, with a fibers, Shimazdu columns, Combi-Pal oven rampAOC-5000, and all plus other auto-sampler. method parameters A 2 cm long are Stablefelx the same DVBas Tomasino/CAR/PDMS et al. combination [22]. Samples SPME were fiber analyzed (50/30 onµm a thickness,Shimazdu 24QP2010 gauge, GC Supleco,MS (Shimazdu, Bellefonte, Columbia PA) was, usedMD, forUSA HS-SPME.) equipped Prior with to a use,Shimazdu the SPME Combi fiber- Palwas AOC conditioned-5000 plus at auto 250 ◦-Csampler. in the injection A 2 cm long port Stablefelx for 1 h. Prior DVB/CAR/PDMS to each sample combination analysis, the SPME SPME fiber fiber (50/30was further µm thickness, conditioned 24 gauge, in an NDL Supleco heater, Bellefonte, attached to PA the) was Combo-Pal used for autosampler HS-SPME. inPrior nitrogen to use for, the 10 SPMminE at fiber 250 ◦ wasC. conditioned at 250 °C in the injection port for 1 h. Prior to each sample analysis, the SPMEThe fiber chromatography was further conditioned configuration in containedan NDL heater dual columns attached connected to the Combo in series-Pal using autosampler a deactivated in nitrogenuniversal for press-tight 10 min at connector250 °C. (Restek, Bellefonte, PA, USA); a Stabilwax column (30 m 0.25 mm × ID The0.5 µ chromam film thickness,tography polyethylene configuration glycol, contained Restek) dual connected columns to a connected Rxi-1MS column in series (15 using m 0.25 a × × deactivatedmm ID 0.5 universalµm film thickness,press-tight 100% connector dimethyl (Restek, polysiloxane, Bellefonte, Restek). PA, USA); The a GC Stabilwax used helium column as (30 the × mcarrier × 0.25 gas mm set ID × 0.5 at a linear μm film velocity thickness, of 33.5 polyethylene cm/s. The GC gl ovenycol, temperatureRestek) connected was held to a atRxi 35-1MS◦C for column 3 min, (15then m × 0.25 ramped mm up to ID × 0.5 μm 250 ◦C at 4 film◦C/min, thickness, and then 100% held dimethyl at this temperature polysiloxane, for 10 Restek). min. The The interface GC used and helium as the carrier gas set at a linear velocity of 33.5 cm/s. The GC oven temperature was held at

Foods 2020, 9, 802 5 of 14

MS source temperature was set at 250 ◦C and 200 ◦C, respectively, with the MS source operated in electron impact (EI) mode at an ionisation energy of 70 eV. For aroma analysis, we used peak areas integrating using the main target ion (m/z) for each compound (Table S1). Compounds were identified by matching their mass spectra to the NIST11 mass spectra library (National Institute of Standards and Technology) and comparison to chemical standards (Table S1). Chromatogram analysis was conducted using GCMSsolutions version 4.20 (Shimazdu, USA).

2.3. Sensory Analysis

After five months of bottle aging at 13 ◦C, sensory analysis was conducted using a sensory panel composed of 17 winemakers from the Willamette Valley, Oregon. The age range of panelists was 25 to 66 and each winemaker had a minimum of 5 years’ experience producing white wine. Panelists were screened for oral lesions, specific anosmia, and cigarette use. A positive response for any of the questions resulted in exclusion. Each panelist tasted 10 wines presented in a random order using an incomplete block design, which included the 8 treatments listed and two randomly designated replicate samples in one two-hour session. Wine glasses were labeled with randomly generated three-digit identifiers. Any background odors were eliminated with air purifiers and the temperature of the room was kept at 20 2 C. ± ◦ This experiment utilized Napping® followed by Ultra-flash-profiling (UFP) [23,24]. In brief, sketch paper (50 lb., 45.7 cm 61 cm) and pens were placed in front of the panelist. Panelists were × asked to refrain from smelling the wine samples as mouth feel analysis was the main objective of the sensory tests. They were instructed to immediately take the sample into their mouth. Tasters grouped the wines based on similarity of mouthfeel, with wines placed closer on the paper to wines of similar mouthfeel and wines which were very different in mouthfeel being placed further apart. Once the wines were placed on the paper, each panelist was asked to enrich the wine(s) with descriptors related to mouth feel which would characterize the differences between wines written near the wine/group (UFP). UFP terms were combined when obvious synonyms were utilized by panelists. This study utilized an incomplete block design for replication, where each panelist received two replicate samples per tasting, which resulted in a complete replication of each treatment across all panelists.

2.4. Data Analysis Analysis of variance was used to interpret the chemical parameters with treatment types using R studio version 3.2.1 (R consortium, Boston, MA, USA). Tukey’s HSD test and 95% confidence intervals were utilized to assess the impact of winemaking treatment on alcohol concentration, acetic acid concentration, malic acid degradation, as well as time to complete MLF. Aroma composition was analyzed using principal component analysis. Aroma composition and sensory data analysis was conducted using XLSTAT (Addingsoft co., New York, NY, USA) and the FactoMineR package from R version 3.2.1 [25]. Napping® data were obtained using a tape measure (millimeters) from the left (X) and bottom edges (Y) relative to the original orientation of the paper to the panelists. These measurements were utilized to generate Multiple Factor Analysis. Correspondence analysis was used to evaluate the UFP terms.

3. Results All treatments completed alcoholic fermentation within 35 days, although treatments where T. delbrueckii Prelude™ was added were initially slower to metabolize glucose/fructose (Figure2). MLF completed in 8 days in all concurrently inoculated treatments (malic acid <0.5 mg/L), while it took between four and five weeks to complete in sequential inoculations (Figure3). When combining the length of time for the completion of both the alcoholic and malolactic fermentation, there were significant differences (p < 0.05) between the treatments. Chardonnay wines produced with a concurrent inoculation strategy completed the fermentations in 26 days, while those produced using a sequential Foods 2020 9 , , 802 Foods 2020, 9, x FOR PEER REVIEW 6 of 14 6 of 15

concurrent inoculation strategy completed the fermentations in 26 days, while those produced using fermentation strategy took between 62 and 82 days to complete (Table1). Sequential fermentations a sequential fermentation strategy took between 62 and 82 days to complete (Table 1). Sequential where T. delbrueckii had been added pre-AF contained lower concentrations of malic acid at the end of fermentations where T. delbrueckii had been added pre-AF contained lower concentrations of malic AF than wines where T. delbrueckii had not been added (Figure3). Wines were assessed for a number acid at the end of AF than wines where T. delbrueckii had not been added (Figure 3). Wines were of parameters after the completion of AF and MLF (Table1). While there was no significant di fference assessed for a number of parameters after the completion of AF and MLF (Table 1). While there was in acetic acid concentration between the fermentation treatments conducted at the same temperature no significant difference in acetic acid concentration between the fermentation treatments conducted (Table1), there was a significant di fference (p < 0.05) in acetic acid between wines fermented at different at the same temperature (Table 1), there was a significant difference (p < 0.05) in acetic acid between temperatures. Ferments conducted at 15 ◦C contained significantly higher concentrations of acetic acid wines fermented at different temperatures. Ferments conducted at 15 °C contained significantly compared to ferments conducted at 21 ◦C (Table1). higher concentrations of acetic acid compared to ferments conducted at 21 °C (Table 1).

280 240 15 °C 200 160 120

80 Glu+Fru (mg/L) Glu+Fru 40 0 280 240 21 oC 200 160 120 80

Glu+Fru (mg/L) Glu+Fru 40 0 0 7 14 21 28 35 42 49 Time (days)

Figure 2. ChangesFigure in glucose 2. Changes and fructose in glucoseduring alcoholic and fructose fermentation during atalcoholic either 15 or fermentation 21 ◦C. Concurrent at either 15 or 21 °C. AF and MLF conductedConcurrent by S. AF cerevisiae and MLFD47 conducted and O. oeni byBeta S. cerevisiae with pre-fermentation D47 and O. oeni addition Beta with of T. pre-fermentation delbrukeii Prelude™addition(∆); AF of by T.S. delbrukeii cerevisiae D47Prelude™ with pre-fermentation (Δ); AF by S. cerevisiae addition D47 of T. with delbrukeii pre-fermentationPrelude™ addition of T. (♦); AF by S. cerevisiaedelbrukeiiD47 (Prelude™); concurrent (◊); AF AF by and S. MLFcerevisiae conducted D47 (□ by); cS.oncurrent cerevisiae AFD47 and and MLFO. oeniconductedBeta by S. cerevisiae ( ). Values are meansD47 fromand O. triplicate oeni Beta fermentations. (◦). Values are means from triplicate fermentations. ◦

Foods 2020, 9, 802 7 of 14 Foods 2020, 9, x FOR PEER REVIEW 7 of 15

2.0

1.6 151°C 1.2 0.8

Malic acid (g/L) acid Malic 0.4

0.0 2.0

1.6

1.2 211°C 0.8

Malic acid (g/L) acid Malic 0.4

0.0 0 7 14 21 28 35 Time (days)

Figure 3.FigureMalic acid3. Malic concentration acid concentration during malolactic during malolactic fermentation fermentation conducted conducted at either at 15 either or 21 15◦C. or 21 °C. ConcurrentCo MLFncurrent and AF MLF conducted and AF by conductedO. oeni Beta by andO. S. oeni cerevisiae Beta andD47 S. with cerevisiae pre-fermentation D47 with additionpre-fermentation of T. delbrukeiiadditionPrelude of T.™ delbrukeii(∆); sequential Prelude™ MLF (Δ by); sO.equential oeni Beta MLF following by O. AFoeni by BetaS. cerevisiaefollowingD47 AF withby S. a cerevisiae T. delbrukeii O. oeni pre-fermentationD47 with addition a pre-fermentation of additionPrelude ™of (T.♦); delbrukeii sequential Prelude™ MLF by (◊); seBetaquential following MLF by AF O. by oeni Beta S. cerevisiae D47 (); concurrent MLF and AF conducted by O. oeni Beta and by S. cerevisiae D47 ( ). following AF by S. cerevisiae D47 (□); concurrent MLF and AF conducted by O. oeni Beta◦ and by S. Values are means from triplicate fermentations. cerevisiae D47 (◦). Values are means from triplicate fermentations. Table 1. Time to completion of alcoholic fermentation (AF) and malolactic fermentation (MLF) of ChardonnayTable wine 1. Time treatments to completion by S. cerevisiae of alcoholicD47 and fermentationO. oeni Beta (A andF) and final malolactic wine and acetic (MLF) of acid concentrations.Chardonnay wine treatments by S. cerevisiae D47 and O. oeni Beta and final wine ethanol and acetic acid concentrations. Winemaking Treatment Days to Complete AF & MLF Alcohol % (v/v) Acetic Acid (g/L) a b a Concurrent AF and MLF at 15 ◦C 26 0 Days to14.14 Complete0.09 AF Alcohol0.72 0.03 % Acetic Acid Winemaking Treatment ± ± ± Concurrent AF and MLF at 15 C + ◦ 26 0 a 13.87& MLF0.06 a 0.72(v/v) 0.03 a (g/L) pre-ferment addition of T. delbrueckii ± ± a ± b a Concurrent AF and MLF at 15 °C b 26 ± 0 d 14.14 ± 0.09 a 0.72 ± 0.03 Sequential AF and MLF at 15 ◦C 68 0 14.64 0.06 0.70 0.01 ± ± ± SequentialConcurrent AF and MLF atAF 15 andC + MLF at 15 °C + pre-ferment ◦ 82 0 c 14.5426 ± 00.08 a c 13.870.71 ± 0.00.016 aa 0.72 ± 0.03 a pre-fermentadd additionition of of T. delbrueckiidelbrueckii ± ± ± a b b Concurrent AF and MLF at 21 ◦C 26 0 14.18 0.10 0.58 0.01 ± ± b ± d a ConcurrentSequential AF and MLF AF at 21andC MLF+ at 15 °C 68 ± 0 14.64 ± 0.06 0.70 ± 0.01 ◦ 26 0 a 13.82 0.04 a 0.59 0.01 b pre-ferment addition of T. delbrueckii Sequential AF and MLF at 15 °C + pre-ferment± ± c ± c a b 82 ± 0 c 14.54 ± 0.08 b 0.71 ± 0.01 Sequential AF and MLF at 21 ◦C 62 0 14.55 0.06 0.56 0.01 addition of T. delbrueckii ± ± ± Sequential AF and MLF at 21 C + ◦ 62 0 b 14.43 0.04a c 0.58 0.00bb b pre-fermentConcurrent addition of T.AF delbrueckii and MLF at 21 °C ± 26 ±± 0 14.18 ± ±0.10 0.58 ± 0.01 a–d Concurrent AF and MLF at 21 °C + pre-ferment Values with different subscripts within each column are significantly di26fferent ± 0 a by p-value <13.820.05 according ± 0.04 a to 0.59 ± 0.01 b Tukey’sadd HSD.ition of T. delbrueckii Sequential AF and MLF at 21 °C 62 ± 0 b 14.55 ± 0.06 c 0.56 ± 0.01 b Sequential AF and MLF at 21 °C + pre-ferment There were also significant differences (p < 0.05) in the final ethanol62 ± 0 b concentrations14.43 ± 0.0 of4 the c wines.0.58 ± 0.00 b addition of T. delbrueckii For fermentations conducted at 15 ◦C, there were significant differences (p < 0.05) between all treatments a–d Values with different subscripts within each column are significantly different by p-value <0.05 for ethanol concentration (Table1). Wines produced by concurrent AF and MLF plus T. delbruekii according to Tukey’s HSD. PreludeTM addition pre-fermentation had the lowest ethanol, while ferments conducted by sequential fermentationThere contained were thealso highest significant ethanol differences content. (p At < 0 21.05)◦C, in both the final the concurrent ethanol concentrations inoculated wines of the wines. For fermentations conducted at 15 °C, there were significant differences (p < 0.05) between all treatments for ethanol concentration (Table 1). Wines produced by concurrent AF and MLF plus T.

Foods 2020, 9, 802 8 of 14 contained lower alcohol than wines produced by sequential inoculation (Table1). Overall, the highest ethanol concentration was measured in wines fermented at 15 ◦C where MLF occurred after alcoholic fermentation, while the lowest was in wines fermented at 21 ◦C where Prelude™ had been inoculated and MLF occurred concurrently. Principle component analysis of the aroma compounds showed a clear separation between sequentially inoculated wines and concurrent inoculated wines along PC1 and PC2 (Figure4). A total of 77% of the variance is explained by the first three principal components. Within the sequential inoculations, there was a clear separation by temperature, which was not found in the concurrent inoculated wines. The influence of Torulaspora delbrueckii on aroma composition was seen by wines in the positive F3 direction and wines without Torulaspora delbrueckii in the negative F3 direction (Figure4). Napping® yielded broadly defined groupings without obvious consistency between temperature or treatment type (Figure5). While it did appear that wine treatment correlated with di fferences in mouthfeel, the differences did not appear consistent between temperatures. Correspondence analysis utilizing the UFP data showed how descriptors were associated with each wine (Figure6). As seen in Figure6, three of the four co-inoculated treatments are di fferentiated by groupings on the negative F2 axis; three of the four T. delbrueckii treatments also lie on the negative F2 axis. The wines appear to vary in their degree of difference in Napping® location and UFP data. Foods 2020, 9, x FOR PEER REVIEW 9 of 15 Foods 2020, 9, 802 9 of 14

Figure 4. vector loading and means for Sequential (Seq) and Concurrent (Co) treatments at two different temperatures (15 ◦C and 21 ◦C) and with Figureinoculation 4. Aroma of T.compound delbrueckii vector(P). (A loading) Contains and PC1 means and PC2for Sequential and (B) contains (Seq) and PC2 Co andncurrent PC3. (Co) treatments at two different temperatures (15 °C and 21 °C) and with inoculation of T. delbrueckii (P). (A) Contains PC1 and PC2 and (B) contains PC2 and PC3.

Foods 2020, 9, 802 10 of 14 Foods 2020, 9, x FOR PEER REVIEW 10 of 15

5 Foods 2020, 9, x FOR PEER REVIEW Co21 10 of 15 4 5 Co21 3 4

2 3 CoP15 2 1 CoP15 Seq21 1 CoP21 0 Seq21 CoP21

F2 (21.98 (21.98 F2 %) 0 SeqP15

F2 (21.98 (21.98 F2 %) SeqP15 -1 -1

-2 -2 SeqP21SeqP21 Co15Co15

-3 -3 Seq15 -4 Seq15 -4 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 -6 -5 -4 -3 -2 -1F1 (25.650 %)1 2 3 4 5 6 F1 (25.65 %) Figure 5. Multiple Factor Analysis of Napping® data for mouthfeel for Sequential (Seq) and Figure 5. MultipleConcurrent Factor (Co) treatments Analysis at of two Napping different® temperaturesdata for mouthfeel (15 °C and for21 °C) Sequential and with inoculation (Seq) and of Concurrent T. delbrueckii (P). ® Figure(Co) treatments 5. Multiple at two Factor different Analysis temperatures of Napping (15 ◦C anddata 21 ◦ forC) and mouthfeel with inoculation for Sequential of T. delbrueckii (Seq) and(P). Concurrent (Co) treatments at two different temperatures (15 °C and 21 °C) and with inoculation of 1.5 T. delbrueckii (P).

1.5 1 low mouthfeel high mouthfeel

0.5 round 1 SeqP21 acidic Co21 smooth high mouthfeelunbalanced low mouthfeel bitter rich CoP15 0 Seq21 astringent sweetness dry 0.5 roundCo15 thin F2 (19.90 %) (19.90 F2 SeqP21 balanced acidic SeqP15Co21 -0.5 smooth Seq15 unbalanced prickly CoP15CoP21 bitter rich Seq21salty astringent 0 sting -1 sweetness dry Co15 thin F2 (19.90 %) (19.90 F2 balanced SeqP15 chewy -0.5 -1.5 Seq15 -1.5 -1prickly -0.5 CoP210 0.5 1 1.5 F1 (25.88 %) salty sting Figure 6. -1Correspondence analysis of ultra-flash profiling (UFP) descriptors for Sequential (Seq) and Figure 6. Correspondence analysis of ultra-flash profiling (UFP) descriptors for Sequential (Seq) and Concurrent (Co) treatments at two different temperatures (15 C and 21 C) and with inoculation of T. Concurrent (Co) treatments at two different temperatures (15 °C◦ and 21 °C)◦ and with inoculation of delbrueckii (P). T. delbrueckii (P). chewy

4. Discussion-1.5 -1.5 -1 -0.5 0 0.5 1 1.5 Concurrent inoculation of AF and MLF during Chardonnay wine production was explored in F1 (25.88 %) the present study. While there have been contradictory reports in literature regarding the benefits of this technique regarding fermentation kinetics and sensory impact [26–28], the results from the presentFigure study 6. Correspondence support the use analysis of concurrent of ultra-fla inoculationsh profiling as (UFP) a method descriptors to significantly for Sequential reduce (Seq) the and length Concurrent (Co) treatments at two different temperatures (15 °C and 21 °C) and with inoculation of T. delbrueckii (P).

Foods 2020, 9, 802 11 of 14 of AF and MLF. This is likely due to the choice of yeast and malolactic bacteria used in the present study as others have noted that the specific yeast and bacteria combination can have a significant influence on the success of fermentation [14,29–31]. In the present study, the difference between when fermentations were completed in the co-inoculated ferments vs. sequential ferments was as large as 56 days. Concurrent inoculated ferments allow for earlier SO2 additions to minimize oxidation and microbial spoilage as well as earlier release of product to the market [15]. The addition of a high population of the non-Saccharomyces yeast T. delbrueckii did not impact the kinetics of the concurrent inoculated AF and MLF. While interactions between non-Saccharomyces yeast and Saccharomyces cerevisiae have been reported previously [32], little is known regarding how these impact O. oeni. Results from this study suggest that high populations of T. delbruekii will not hinder O. oeni conducting MLF, suggesting that concurrent inoculation may still be a viable option in years of high microbial load on grape skins. It would be interesting to perform a concurrent inoculated fermentation where there is a high background population of H. uvarum as this yeast is the most common yeast found on grapes at harvest. Varying the timing of the fermentations (AF and MLF) impacted the aroma composition of the wines, with sequential inoculations also being impacted by fermentation temperature. The choice of yeast and bacteria strains is known to alter aroma composition [12,15] and differences in aroma composition for sequential inoculation compared to concurrent inoculation has also been demonstrated in other wines [12,33]. The main aroma compound differences between the co versus sequential inoculated wines are that the sequentially inoculated wines, specifically at the cooler fermentation temperature, were characterized by a greater number of compounds, specifically ethyl esters. The concurrent inoculated wines were characterized by 1-octanol, 1-decanol, ethyl lactate, and butyrolactone. The sequential inoculations were separated based on temperature, with the higher fermentation temperature (21 ◦C) wines characterized by isobutyric acid, 2-methylbutanoic acid, diethyl succinate, , and isopentyl hexanoate. The cooler sequentially inoculated wines were characterized by a greater number of aroma compounds including ethyl esters, gamma-terpinene, and 1-hexanol. These results are in agreement with previous work that has shown that sequential and concurrent inoculations alter the fruity and lactic aromas of wines [34]. Ethyl lactate and diethyl succinate are known to influence the buttery and creamy aspects of wine [35], ethyl esters are known to have fruity aromas [36], and fatty acids are linked to both fruity and dairy aromas [37]. These aroma compounds are most likely influencing aroma perception as fatty acids and ethyl esters are known to be important impact odorants [37,38]. Sequential fermentations alter the metabolites available to the malolactic bacteria compared to the yeast, while concurrent inoculated fermentations would have the same available to both. Therefore, the timing of MLF has the potential to change the aroma characteristics of the wine, as described above, which may or may not be desirable depending on the targeted wine style. In addition, wines produced by concurrent inoculated fermentations will be able to have SO2 added at an earlier time point and reduce the likelihood of aroma changes due to oxidation. For example, in the present study, no SO2 could be added for 60–80 days to wines that underwent sequential fermentations, while SO2 could be added after 26 days if wine was produced by concurrent inoculation. The present study is one of the few to determine mouthfeel differences due to MLF timing and temperature. While mouthfeel differences between the various wines were noted, the differences did not necessarily align with differences in pH, acetic acid, and residual sugar content. Wines also did not group based on temperature of fermentation or timing of the MLF. These findings suggest that while the range of winemaking procedures investigated in this study can affect Chardonnay wine mouthfeel, there was not one dominant factor driving mouthfeel differences. Although the addition of T. delbruekii did not impact fermentation kinetics, it did impact other wine parameters. For example, a drop in malic acid due to the addition of T. delbrueckii was noted. Other non-Saccharomyces yeasts have the ability to partially degrade malic acid [39,40] and based on our results, T. delbruekii also has this trait. T. delbrueckii could be a viable alternative to O. oeni inoculation Foods 2020, 9, 802 12 of 14 for partial degradation of malic acid in cool climates. This is an important consideration for wines with particularly high levels of malic acid, such as sparkling wines, when MLF can be challenging for O. oeni to complete. A combination of a non-Saccharomyces yeast that can partially degrade the malic acid and O. oeni may be helpful in these situations. It could also be a tool for winemakers who do not wish for a complete MLF as T. delbrueckii cannot completely utilize the malic acid, but can partially degrade it [41]. A reduction in alcohol was also noted when T. delbruekii was added pre-fermentation for the concurrent AF and MLF wines. This reduction in alcohol was not observed if MLF was carried out sequentially. This finding is in alignment with recent studies where an addition of non-Saccharomyecs yeast prior to alcoholic fermentation lowered the final alcohol content of the wine [19]. T. delbrueckii also altered the aroma composition of the wines. Wines made with this yeast were characterized by diethyl succinate, isobutyric acid, isoamyl acteate, acetic acid, hexyl acetate, and 3-methylbutyl octanoate. The change in aroma composition when T. delbrueckii was part of the fermentation was anticipated, as previous work has shown that T. delbrueckii alters aroma composition in a similar fashion as shown in our results [42]. The role of T. delbrueckii on the mouthfeel perception differences between treatments is not completely understood. Three of the four treatments inoculated with T. delbrueckii were consistently grouped together and were all influenced by the descriptive terms balanced, rich, and sweetness. Sequential inoculation and T. delbrueckii at 15 ◦C treatment was grouped separately and was characterized by the terms astringent, unbalanced, thin, and dry. Unlike the concurrent inoculation grouping, this treatment did not have significantly different residual sugar or ethanol, which might explain this discrepancy. Domizio et al. [43] has previously demonstrated the increase in the mannoprotein content of finished wines when T. delbrueckii is inoculated. However, mannoprotein differences alone do not appear to account for the differences between treatments of this study. Understanding the metabolism of T. delbruckii and the secondary metabolites of malic acid degradation could lead to a new understanding of this yeast’s impact on wine sensory evaluation.

5. Conclusions Concurrent inoculated fermentation of Chardonnay significantly reduced the time needed for the completion of both fermentations. No adverse impacts on wine quality were noted for concurrently inoculated ferments as acetic acid increases were driven by lower fermentation temperature rather than MLF timing. Timing and temperature of MLF impacted volatile aroma composition, while mouthfeel was affected to a lesser extent. The addition of a high population of a non-Saccharomyces pre-fermentation did not impact concurrently inoculated MLFs, but did result in significant differences in aroma composition and mouthfeel perception compared to controls. This study has demonstrated the usefulness of concurrent inoculation for a more rapid completion of AF and MLF fermentations as well as how MLF timing and temperature may impact wine qualities such as aroma and mouthfeel.

Supplementary Materials: The following are available online at http://www.mdpi.com/2304-8158/9/6/802/s1, Table S1: Qualification parameters for HS-SPME-GCMS analysis of aroma compounds in Chardonnay wines. Author Contributions: J.O. and E.T. conceived and designed the experiments; A.S. performed the experiments and analyzed the data; Q.P. did the aroma compound analysis; A.S., J.O., and E.T. wrote the paper. All authors have read and agreed to the published version of the manuscript. Funding: The authors thank the Research Institute for their financial support. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Jackson, R. Wine Science Principles and Applications, 2nd ed.; Elsevier Inc.: Amsterdam, NL, USA, 2008. 2. Jaffe, J.; Valentin, D.; Meunier, J.-M.; Siliani, A.; Bertuccioli, M.; Le Fur, Y.The Chardonnay Wine Olfactory Concept Revisited: A Stable Core of Volatile Compounds, and Fuzzy Boundaries. Food Res. Int. 2011, 44, 456–464. [CrossRef] 3. Robinson, J.; Harding, J.; Vouillamoz, J. Wine Grapes; Ecco: New York, NY, USA, 2012. Foods 2020, 9, 802 13 of 14

4. Cottrell, T.H.E.; Lellan, M.R.M. The Effect of Fermentation Temperature on Chemical and Sensory Characteristics of Wines from Seven White Grape Cultivars Grown in New York State. Am. J. Enol. Vitic. 1986, 37, 190–194. 5. Cutler, L. Focus Chardonnay. In Wine Business Monthly; Wine Communications Group Inc.: Sonoma, CA, USA, 2012. 6. Silver, J.; Leighton, T. Control of Malolactic Fermentation in Wine. 2. Isolation and Characterization of a New Malolactic Organism. Am. J. Enol. Vitic. 1981, 32, 64–72. 7. Gambetta, J.M.; Bastian, S.E.P.; Cozzolino, D.; Jeffery, D.W. Factors Influencing the Aroma Composition of Chardonnay Wines. J. Agric. Food Chem. 2014, 62, 6512–6534. [CrossRef][PubMed] 8. Avedovech, R.M.; Mcdaniel, M.R.; Watson, B.T.; Sandine, W.E. An Evaluation of Combinations of Wine Yeast and Leuconostoc oenos Strains in Malolactic Fermentation of Chardonnay Wine. Am. J. Enol. Vitic. 1992, 43, 253–260. 9. Guzzon, R.; Villega, T.; Pedron, M.; Malacarne, M.; Nicolini, G.; Larcher, R. Simultaneous Yeast–Bacteria Inoculum. A Feasible Solution for the Management of Oenological Fermentation in red must with Low Nitrogen Content—Springer. Ann. Microbiol. 2012, 63, 805–808. [CrossRef] 10. Mills, D.A.; Rawsthorne, H.; Parker, C.; Tamir, D.; Makarova, K. Genomic Analysis of Oenococcus oeni PSU-1 and Its Relevance to Winemaking. FEMS Microbiol. Rev. 2005, 29, 465–475. 11. Wibowo, D.; Eschenbruch, R.; Davis, C.R.; Fleet, G.H.; Lee, T.H. Occurrence and Growth of Lactic acid Bacteria in Wine: A Review. Am. J. Enol. Vitic. 1985, 36, 302–313. 12. Abrahamse, C.; Bartowsky, E. Timing of Malolactic Fermentation Inoculation in Shiraz Grape Must and Wine: Influence on Chemical Composition. World J. Microbiol. Biotechnol. 2012, 28, 255–265. [CrossRef] 13. Rossouw, D.; Du Toit, M.; Bauer, F.F. The Impact of Co-Inoculation with Oenococcus oeni on the Trancriptome of Saccharomyces cerevisiae and on the Flavour-Active Metabolite Profiles during Fermentation in Synthetic must. Food Microbiol. 2012, 29, 121–131. [CrossRef] 14. Muñoz, V.; Beccaria, B.; Abreo, E. Simultaneous and Successive Inoculations of Yeasts and Lactic Acid Bacteria on the Fermentation of an Unsulfited Grape must. Braz. J. Microbiol. 2014, 45, 59–66. [CrossRef] [PubMed] 15. Maarman, B.C. Interaction between Wine Yeast and Malolactic Bacteria and the Impact on Wine Aroma and Flavour. Master’s Thesis, Stellenbosch University, Stellenbosch, South Africa, 2014. 16. Zahavi, T.; Droby, S.; Cohen, L.; Weiss, B.; Ben-Arie, R. Characterization of the Yeast Flora on the Surface of Grape Berries in Israel. 2002, 41, 203–208. 17. Albertin, W.; Miot-Sertier, C.; Bely, M.; Marullo, P.; Coulon, J.; Moine, V.; Masneuf-Pomarede, I. Oenological Prefermentation Practices Strongly Impact Yeast Population Dynamics and Alcoholic Fermentation Kinetics in Chardonnay Grape must. Int. J. Food Microbiol. 2014, 178, 87–97. [CrossRef] 18. Arnink, K.; Henick-Kling, T. Influence of Saccharomyces cerevisiae and Oenococcus oeni Strains on Successful Malolactic Conversion in Wine. Am. J. Enol. Vitic. 2005, 56, 228–237. 19. Contreras, A.; Hidalgo, C.; Schmidt, S.; Henschke, P.A.; Curtin, C.; Varela, C. The Application of Non-Saccharomyces yeast in fermentations with limited aeration as a strategy for the production of wine with Reduced Alcohol Content. Int. J. Food Microbiol. 2015, 205, 7–15. [CrossRef][PubMed] 20. Medina, K.; Boido, E.; Fariña, L.; Gioia, O.; Gomez, M.E.; Barquet, M.; Gaggero, C.; Dellacassa, E.; Carrau, F. Increased Flavour Diversity of Chardonnay Wines by Spontaneous Fermentation and Co-Fermentation with Hansen. Vineae. Food Chem. 2013, 141, 2513–2521. [CrossRef][PubMed] 21. Englezos, V.; Cachón, D.C.; Rantsiou, K.; Blanco, P.; Petrozziello, M.; Pollon, M.; Giacosa, S.; Segade, S.R.; Rolle, L.; Cocolin, L. Effect of Mixed Species Alcoholic Fermentation on Growth and Malolactic Activity of Lactic acid Bacteria. Appl. Microbiol. Biotechnol. 2019, 103, 7687–7702. [CrossRef][PubMed] 22. Tomasino, E.; Harrison, R.; Breitmeyer, J.; Sedcole, R.; Sherlock, R.; Frost, A. Aroma Composition of 2-Year-Old New Zealand Wine and Its Relationship to Sensory Characteristics Using Canonical Correlation Analysis and Addition/Omission Tests. Aust. J. Grape Wine Res. 2015, 21, 376–388. [CrossRef] 23. Pagès, J. Collection and Analysis of Perceived Product Inter-Distances Using Multiple Factor Analysis: Application to the Study of 10 White Wines from the . Food Qual. Prefer. 2005, 16, 642–649. [CrossRef] 24. Reinbach, H.C.; Giacalone, D.; Ribeiro, L.M.; Bredie, W.L.P.; Frøst, M.B. Comparison of Three Sensory Profiling Methods Based on Consumer Perception: CATA, CATA with Intensity and NAPPING®. Food Qual. Prefer. 2014, 32, 160–166. [CrossRef] Foods 2020, 9, 802 14 of 14

25. Lê, S.; Josse, J.; Husson, F. FactoMineR: An R Package for Multivariate Analysis. J. Stat. Softw. 2008, 25, 1–18. [CrossRef] 26. Davis, C.R.; Wibowo, D.; Eschenbruch, R.; Lee, T.H.; Fleet, G.H. Practical Implications of Malolactic Fermentation: A Review. Am. J. Enol. Vitic. 1985, 36, 290–301. 27. Edwards, C.G.; Reynolds, A.G.; Rodriguez, A.V.; Semon, M.J.; Mills, J.M. Implication of Acetic Acid in the Induction of Slow/Stuck Grape Juice Fermentations and Inhibition of Yeast by sp. Am. J. Enol. Vitic. 1999, 50, 204–210. 28. Jussier, D.; Morneau, A.D.; de Orduña, R.M. Effect of Simultaneous Inoculation with Yeast and Bacteria on Fermentation Kinetics and Key Wine Parameters of Cool-Climate Chardonnay. Appl. Environ. Microbiol. 2006, 72, 221–227. [CrossRef] 29. Beelman, R.; Keen, R.; Banner, M.; King, S. Interactions between Wine Yeast and Malolactic Bacteria Under Wine Conditions. Dev. Ind. Microbiol. 1982, 23, 107–121. 30. Henick-Kling, T.; Park, Y.H. Considerations for the Use of Yeast and Bacterial Starter Cultures: SO2 and Timing of Inoculation. Am. J. Enol. Vitic. 1994, 45, 464–469. 31. Osborne, J.P.; Edwards, C.G. Inhibition of Malolactic Fermentation by a Peptide Produced by Saccharomyces cerevisiae during Alcoholic Fermentation. Int. J. Food Microbiol. 2007, 118, 27–34. [CrossRef] 32. Fleet, G.H. Wine Yeasts for the Future. FEMS Yeast Res. 2008, 8, 979–995. [CrossRef] 33. Knoll, C.; Fritsch, S.; Schnell, S.; Grossman, M.; Krieger-Weber, S.; du Toit, M.; Rauhut, D. Impact of Different Malolactic Fermentation Inoculation Scenarios on Riseling Wine Aroma. World J. Microbiol. Biotechnol. 2012, 28, 143–153. [CrossRef] 34. Antalick, G.; Perello, M.C.; de Revel, G. Co-Inoculation with Yeast and LAB under Conditions: Modification of the Aromatic Profile of Wines. S. Afr. J. Enol. Vitic. 2013, 34, 223–232. [CrossRef] 35. Issa-Issa, H.; Guclu, G.; Noguera-Artiaga, L.; López-Lluch, D.; Poveda, R.; Kelebek, H.; Selli, S.; Carbonell-Barrachina, Á.A. Aroma-Active Compounds, Sensory Profile and Phenolic Composition of Fondillón. Food Chem. 2020, 316, 126353. [CrossRef][PubMed] 36. Chisholm, M.G.; Guiher, L.A.; Zaczkiewicz, S.M. Aroma Characteristics of aged Wine. Am. J. Enol. Vitic. 1995, 46, 56–62. 37. Rocha, S.M.; Rodrigues, F.; Coutinho, P.; Delgadillo, I.; Coimbra, M.A. Volatile Composition of Baga : Assessment of the Identification of the would-be Impact Odourants. Anal. Chim. Acta 2004, 513, 257–262. [CrossRef] 38. López, R.; Ortín, N.; Pérez-Trujillo, J.P.; Cacho, J.; Ferreira, V. Impact Odorants of Different Young White Wines from the Canary Islands. J. Agric. Food Chem. 2003, 51, 3419–3425. [CrossRef][PubMed] 39. Ciani, M.; Comitini, F. Non-Saccharomyces Wine Yeasts have a Promising Role in Biotechnological Approaches to Winemaking. Ann. Microbiol. 2010, 61, 25–32. [CrossRef] 40. Benito, Á.; Calderón, F.; Palomero, F.; Benito, S. Combine Use of Selected Schizosaccharomyces pombe and Lachancea thermotolerans Yeast Strains as an Alternative to the Traditional Malolactic Fermentation in Red Wine Production. Molecules 2015, 20, 9510–9523. [CrossRef] 41. Belda, I.; Navascués, E.; Marquina, D.; Santos, A.; Calderon, F.; Benito, S. Dynamic Analysis of Physiological Properties of Torulaspora delbrueckii in Wine Fermentations and Its Incidence on Wine Quality. Appl. Microbiol. Biotechnol. 2014, 99, 1911–1922. [CrossRef] 42. Azzolini, M.; Tosi, E.; Lorenzini, M.; Finato, F.; Zapparoli, G. Contribution to the Aroma of White Wines by Controlled Torulaspora delbrueckii Cultures in Association with Saccharomyces cerevisiae. World J. Microbiol. Biotechnol. 2014, 31, 277–293. [CrossRef] 43. Domizio, P.; Liu, Y.; Bisson, L.F.; Barile, D. Use of Non-Saccharomyces Wine Yeasts as Novel Sources of Mannoproteins in Wine. Food Microbiol. 2014, 43, 5–15. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).